Efficiency Evaluation of Deep Model for Person Re-identification

In this paper, we evaluate the efficiency in training deep models for person re-identification (Re-ID) based on different experimental settings including the number of GPUs and the batch size. To this end, we employ the baseline and PCB to conduct amounts of experiments on Market-1501. The experimen...

Full description

Saved in:
Bibliographic Details
Published inArtificial Intelligence in China Vol. 653; pp. 130 - 136
Main Authors Zhang, Haijia, Wang, Sen, Wang, Nuoran, Liu, Shuang, Zhang, Zhong
Format Book Chapter
LanguageEnglish
Published Singapore Springer Singapore Pte. Limited 2021
Springer Singapore
SeriesLecture Notes in Electrical Engineering
Subjects
Convolutional neural network
Efficiency evaluation
Person re-identification
Online AccessGet full text

Cover

More Information
Summary:In this paper, we evaluate the efficiency in training deep models for person re-identification (Re-ID) based on different experimental settings including the number of GPUs and the batch size. To this end, we employ the baseline and PCB to conduct amounts of experiments on Market-1501. The experimental results indicate that what experimental settings have important effects on the efficiency in training deep models.
ISBN:9789811585982
9811585989
ISSN:1876-1100
1876-1119
DOI:10.1007/978-981-15-8599-9_16